The High Energy Density Scientific Instrument at the European XFEL

Abstract

The European XFEL delivers up to 27000 intense (>10¹² photons) pulses per second, of ultrashort (≤50 fs) and transversely coherent X-ray radiation, at a maximum repetition rate of 4.5 MHz. Its unique X-ray beam parameters enable groundbreaking experiments in matter at extreme conditions at the High Energy Density (HED) scientific instrument. The performance of the HED instrument during its first two years of operation, its scientific remit, as well as ongoing installations towards full operation are presented. Scientific goals of HED include the investigation of extreme states of matter created by intense laser pulses, diamond anvil cells, or pulsed magnets, and ultrafast X-ray methods that allow their diagnosis using self-amplified spontaneous emission between 5 and 25 keV, coupled with X-ray monochromators and optional seeded beam operation. The HED instrument provides two target chambers, X-ray spectrometers for emission and scattering, X-ray detectors, and a timing tool to correct for residual timing jitter between laser and X-ray pulses.

Keywords: X-ray free-electron lasers; high energy density; high-pressure science; relativistic laser–matter interaction; warm dense matter.

Figure 1

Unique bunch pattern of European XFEL. Within an up to 600 µs-long window, between a single and 2700 X-ray pulses can be created. The facility can deliver pulses at difference repetition rates of 4.51, 2.26, or 1.13 MHz, resulting in a pulse spacing of 222, 443 or 886 ns, respectively.

Figure 2

Current peak performance of the SASE2 FEL at photon energies between 6 and 18 keV. The data points represent peak pulse energies measured in 2019 and 2020 (https://xfel.desy.de/operation/performance/; see also Maltezopoulos et al., 2019 ▸). The linac was operated at 11.5 GeV (blue crosses), 14 GeV (black rectangles) and 16.5 GeV (red circles) electron energy.

Figure 3

HED optics and experiment hutch layout. The X-rays enter from the right. Optics hutch: PSLIT – water-cooled four-blade power slits, BIU – beam imaging unit, PAM – photon arrival monitor for X-ray-optical laser timing, ATT – solid attenuator foils, CRL – compound refractive lenses made of Be (CRL3), IPM1 – intensity and position monitor. Experimental hutch: ALAS: incoupling for alignment laser, DPS – differential pumping, CSLIT – clean-up slits, two four-blade assemblies for soft and hard X-rays, respectively, IPM2 – intensity and position monitor, IC1 and IC2 – interaction chambers, AGIPD – MHz repetition compatible X-ray detector, DBENCH – detector bench, SPECTRO – position of downstream spectrometers, IBS – instrument beam stop. The distances on the top are given from the source (center of last undulator segment).

Figure 4

Schematics showing the position of the ReLaX and DiPOLE 100-X laser in a room above the experiment hutch with the interaction chambers IC1,2 (walls not shown).

Figure 5

Beam transport in the XTD1 and XTD6 underground tunnels up to the XFEL Headquarters EXPeriment hall 1 (XHEXP1). The X-rays enter from the right. Not all components are shown. A detailed list of devices and distances can be found in Appendix B .

Figure 6

Calculated total external reflectivity of the combined mirrors M1–M3 for both reflecting coatings plotted on a log scale, inset on linear scale. Shown is the minimum energy cut-off using the B₄C coating for all three mirrors and mirrors M1 and M2 at 3.6 mrad suitable to reduce third harmonic content, and the maximum energy cut-off using the Pt coating for all three mirrors and mirrors M1 and M2 at 1.7 mrad. Also shown is the typically used working point at 2 mrad using the B₄C coating. In all cases the mirror M3 is set to its working point of 1.3 mrad.

Figure 7

Characterization of the best focus of CRL3 and an intermediate focus scheme with CRL1 at 6 keV photon energy. The three false-color insets are single-shot imprints into a LiF crystal at 0.5% beamline transmission.

Figure 8

Characterization of the best focus of CRL3 after intermediate focus with CRL1 at 17.8 keV with a wire scan. The scans took a few minutes with 5% transmission. The y-axis shows in red the transmission (normalized to the incident pulse energy, and with a moving average), as well as its derivative in blue.

Figure 9

Examples of single-pulse spectra. Spectra as recorded by a 2D detector imaging a scintillator screen (top) and their corresponding lineouts (bottom). Left: HED-flex at 6 keV using a Si(111) crystal and lineouts of three different pulses to illustrate the shot-to-shot fluctuation. Right: HIREX-II in the XTD6 tunnel at 17.5 keV using C(110).

Figure 10

Measured energy resolution of the Si(111) monochromator at E = 6055 eV. The transmitted SASE X-ray pulse energy of the four-bounce setup is shown as a function of the Bragg angle (scaled to photon energy) of the second pair of crystals while the first pair was fixed. The Gaussian fit yields a FWHM of ΔE = 0.72 eV and ΔE/E ≃ 1.2 × 10⁻⁴.

Figure 11

View into the HED laser bay with both HIBEF high-power lasers DIPOLE 100-X (left) and ReLaX (right).

Figure 12

Typical images of PAM. (a) Spatial encoding: vertical versus horizontal position; time mapping results form an angle between the laser and the X-ray on the sample (2 µm-thick Si₃N₄). The temporal window (horizontal) is about 3.6 ps. (b) Spectral encoding image: lateral position versus the wavelength mapped to time by a chirped laser pulse collinear with the X-ray perpendicular to the sample (100 µm-thick YAG:Ce). The temporal window (horizontal axis of the image) is about 1.7 ps. Both show a shadow due to the X-ray-induced opacity change of the sample. At t ₀ both beams are timed. The horizontal edges in (a) result from the upstream partially closed power slits which are not influencing (b).

Figure 13

Floor plan of HED experiment hutch with the positions of IC1 (right, rectangular), IC2 (center, round) and the AGIPD on the detector bench on the left. The XFEL beam enters from the right. The dark-gray areas indicate hutch infrastructure, while the orange areas mark access paths. Dimensions are given in millimetres.

Figure 14

(a) Cut through the schematic of the IC1 target chamber showing the vertical breadboard with circular rails. (b) Top view of the interior of IC1; the access doors are at the bottom. In red, a typical beam path for the ReLaX laser is shown, and the back line indicates a sample viewing system. The beam transport of the DiPOLE-100X high-energy laser in IC1 is flexible and customized configurations are currently evaluated.

Figure 15

EUCALL standard frame design available at HED. The outer (blue) frame is unique to HED. The inner-frames (in gray/silver) can be customized for each experiment. This standard is employed at other beamlines at the EuXFEL, as well as other facilities such as ESRF and ELI beamlines.

Figure 16

Kα fluorescence of a 5 µm Cr foil irradiated by an 10 µm X-ray FEL spot for calibration purposes (black-dashed) with fit (red). The spectral broadening of the lines by the 80 mm radius-of-curvature and 40 µm-thick HAPG crystal is ∼2 eV. Reproduced from Preston et al. (2020 ▸).

Figure 17

3D schematics of up- and downstream HAPG spectrometers and three diced analyzers mounted on the curved rail system in IC1, leaving the horizontal breadboard free for laser optics and further instrumentation.

Figure 18

The spectra of the Si(533)-monochromated beam scattered from PMMA (elastic, blue) and single crystal diamond (inelastic, red) resolved with a diced Si (533) analyzer crystal. For better comparison, the elastic signal is reduced by a factor of ten. For diamond, the peaks correspond to phonon creation and annihilation. The spectrum is reproduced from Wollenweber et al. (2021 ▸).

Figure 19

Model of the Jungfrau detector in an air-box, developed at the HED Instrument for in-vacuum operation.

Figure 20

Interaction chamber IC2. (a) Vacuum chamber. (b) Platform for high-pressure research with diamond anvil cells.

Figure 21

Dynamic laser compression platform optimized for flexible experimental geometries. For the drive laser, this is realized through five individual laser entrance ports at angles of 22.5°, 45°, 67.5°, 90°, and 112.5° with respect to the incident X-ray beam, to which the DiPOLE beam transport is coupled via an external periscope system. The VISAR beam enters the chamber through a single port and covers an angular range from 202.5° to 310° via a pair of fixed and mobile mirrors inside the chamber. The Varex detector system can be rotated around the interaction point in steps of 7.5°.

Figure 22

(a) HIBEF AGIPD 1M detector and coverage as a function of X-ray energy. (b) Varex twin detector system and coverage as a function of X-ray energy for the downstream position.

Figure 23

The changes of X-ray diffraction patterns of iron and nitrogen during in situ chemical reaction by XFEL pump-and-probe. An Fe foil surrounded by N₂ was pre-compressed to 5 GPa in a DAC. The percentages in run A indicate the transmission (fluence) of the XFEL. At 100% XFEL transmission, run B contains 20 consecutive pulses while run C has consecutive pulses over 11 s. For more details and figure credit, refer to Hwang et al. (2021 ▸).

Figure 24

HED focusing optics. The X-rays enter from the left. The coordinate system starts at the center of U33. On the right, in red, are the achievable best focus conditions for 5–25 keV for monochromatic (seeded) beam.